The etiology of spondylolysis is considered to be a stress fracture at the pars interarticularis. To understand the stress distributions within the spinal elements of a motion segment or functional spinal unit (FSU), we developed a finite element model of the ligamentous L3-S1 segment. The details, including element properties, nonlinear behavior, material properties, and experimental validation, are reported elsewhere (3,4,5,9,10,11,12). Stress distributions in various structures were analyzed in flexion, extension, lateral bending, and axial rotation in response to 400 N axial compression and 10.6 Nm moment. Von Mises stresses at the various structures were computed. The predicted highest stresses were found in the pars interarticularis. Furthermore, the stresses in extension and rotation were higher than in flexion and bending (Fig. 3.1). These results support the clinical observation that pars fracture most often occurs in extension and rotation (13,14).

For diagnosing spondylolysis, usually plain radiographs, computed tomography (CT) scans, and scintigrams are used. Spondylolysis can be classified into three stages: early, progressive, and terminal (15). The early stage, which follows the stress fracture, shows as a hairline on CT scans. At the progressive stage, the gap at the defect site is clearly visible. The terminal stage is same as pseudoarthrosis, showing sclerosis at the fracture site. Even using CT scan, the early stage is sometimes very difficult to diagnose. The Scottish
terrier collar sign on oblique radiographs is the famous and reliable sign. However, the reliability of the sign is questionable in detecting early stages of this disorder.

Magnetic resonance imaging (MRI) is less invasive than computed tomography or plain radiography, and more recently, we have shown that a high signal at the pedicle in an axial slice of T2-weighted MR images can be used effectively for detecting the early stage of the disorder (4). Figure 3.2 demonstrates CT and MRI scans of early stage spondylolysis in two pediatric patients. Both patients show early stage spondylolysis on CT, and the adjoining pedicles show high signal (arrows) on MRI T2-weighted image.

We investigated the pathomechanism of the signal change of pedicle using finite element (FE) analysis. The stresses at the pedicle were second highest, next to the pars interarticularis. Thus, the highest stress at the pars may induce a stress fracture at the site,
while the second highest stress may cause bone marrow edema due to stress reaction like a condition of bone bruise.

Using MRI, other pediatric lumbar disorders, such as apophyseal bony ring fracture and herniated nucleus pulposus (HNP), can also be diagnosed. Thus, this method can be a valuable tool for screening pediatric patients complaining of low back pain. We propose an axial slice (D) through the pedicle while taking slices around the disc space (A-C) on MRIs (Fig. 3.3). Those slices help diagnosis the apophyseal ring tear and the HNP, while the axial slice facilitates early diagnosis of lumbar spondylolysis. Thus, the addition of the extra MRI slice will assist in correct diagnosis of the three major lumbar disorders in pediatric patients who visit the clinic complaining of low back pain.

Spondylolysis in its early and progressive stages responds to conservative treatment. Terminal stage defects (pseudoarthrosis) mandate surgical interventions to promote bony union. For conservative treatment, a brace to immobilize the fracture site is usually prescribed along with the cessation of sport activities for at least 3 months (5,15). Finite element analysis was used to determine the relationship between the load type and direction of fracture (Fig. 3.4). This information may help determine the most optimal brace type to prescribe. The stress data indicate that patients can be divided into two categories, extension- or rotation-induced spondylolysis. For patients in whom the fracture indicates a sagittal orientation, the optimal brace would be the one that immobilizes rotary motion. But patients with the fracture angle more coronally oriented need a brace that will effectively immobilize motion in extension. To further prove the hypothesis, clinical prospective trials using a specific brace according to the CT-based fracture angle are essential.